FEATURE
INGENIEUR
Enhancing Smart City Environment using Machine Learning
By Michelle Ng Xin Yi , Yu Hui Ling , Zhang Xu , Ong Kheng Chong , Lang Yiming , Dr Ten Joon Yoon , Ir . Kim Kek Seong , Ir . Assoc . Prof . Dr Parthiban Siwayanan , Ir . Dr Ban Zhen Hong School of Energy and Chemical Engineering , Xiamen University Malaysia
Ir . Prof . Dr Thomas Choong Shean Yaw Faculty of Engineering , Universiti Putra Malaysia
Emerging smart cities may offer a potential strategy to handle the challenges that arise from the increase in urban population . The six components for the development of a smart city are shown in Figure 1 . Currently , there are no set characteristics or requirements that define what constitutes a smart city . Nonetheless , various literature has put forward a variety of definitions . The main idea involves the integration of technologies , such as Artificial Intelligence ( AI ), the Internet of Things ( IoT ), information and communications technology ( ICT ), and Web 2.0 technology , into the critical infrastructure of a city to build an instrumented , interconnected , and intelligent system for improved sustainability and liveability . According to Chourabi et al . ( 2012 ), the installation of instruments enables the capture of real-time data through the use of sensors , meters , personal devices , and other data-acquisition devices . With an interconnected system , the data can then be integrated into a central platform and transmitted through different city services . Intelligent devices utilise the provided information to analyse , adapt , and react to different situations and maximise operational efficiency .
Smart Environment seeks to ensure that the city is free from pollution to promote a healthy and sustainable environment and maintain a high quality of life for its inhabitants . One of the primary challenges for the Smart Environment is air pollution . The rapid expansion of various industries is a major contributor to air pollution and has been met with heavy environmental pressure . Prolonged exposure to high levels of pollutants from industries can harm human health by causing respiratory illnesses like asthma and bronchitis , and increasing the risk of cancers of the lung , mouth , and throat as reported by Gourd ( 2022 ). Moreover , the accidental release of hazardous gas from industrial plants can negatively impact public safety and health . The most infamous example of an accidental release is perhaps the Bhopal disaster , wherein residents near the site were exposed to the highly toxic methyl isocyanate gas , causing numerous deaths and injuries . Hence , there is increasing interest in a better understanding of pollutant and hazardous gas dispersion as well as detection and source localisation methods for pollution sources and toxic gas leaks .
In the past , hazardous gas dispersion studies were conducted through large-scale experimental programmes . These experiments were commonly intended to broaden the understanding of atmospheric dispersion of hazardous gases in the case of accidental releases . The tests typically utilised a system of sensors installed downwind from the release point to measure the concentration of the gas of interest . While these studies provided insight into the dispersion of gases in all stages , starting at the source , to passive dispersion , the measured
26 VOL 93 JANUARY-MARCH 2023